Particle Swarm Optimisation of Electromagnetic Soundings

We discuss through synthetic and real data some the application of PSO in electromagnetic soundings. The suggested approach can be easily adapted to resistivity soundings (RS), time domain soundings (TDEM) , magneto-telluric (MT) and audio-magneto-telluric survey (AMT). We propose an overview on the PSO for solving 1D problems with a priori information and/or lateral constraints. The application of PSO on AMT data is suggested by the high speed of convergence to a problem’s solution respect other evolutionary methods. Application on the synthetic dataset allow us to analyze the relevance of the setting parameters, and to select the optimal solutions when a priori information or additional constraints are introduced. We demonstrate how PSO could be an effective approach in AMT data processing (1D). The results can be selected as starting model for a subsequent gradient-based inversion